Homepage and feeds
A personal homepage or feed for every user, refreshed as they browse.
Recommendation engine
Personal recommendations for every user, in real time.
vecsai Recommend learns from what each user views, buys, watches and skips, and from what the items themselves are about. Recommendations update as the session unfolds, new items can be recommended on the day they are added, and you decide what may and may not be shown.
Because you watched The Long Orbit
- Deep Signal
- Red Horizon
- Quiet Station
- Last Light
New this week for you
- Northbound
- Salt & Iron
- The Archive
- Glasswater
Updated after every playexample
Features
What vecsai Recommend does
Real-time personalisation
Every view, click and purchase updates the next recommendation. No nightly batch job.
Cold start, solved
New items are recommended from their content before they have any interactions. New users get relevant picks from their first clicks.
Many scenarios, one engine
Items for a user, items similar to an item, users for an item, next to watch, frequently bought together.
Business rules
Filter by stock, region or rights; boost margins or new releases; keep pinned positions fixed.
Diversity and freshness
Control how varied recommendations are and how strongly recent content is favoured.
Tested before release
Run a new recommendation strategy through vecsai Sim before your users ever see it.
Use cases
Where teams use it
“You may also like”
Similar and complementary items on product, article and video pages.
Email and push
Personal picks for newsletters and notifications, under the same rules as the site.
FAQ
Questions about vecsai Recommend
What data do you need?
A catalogue of items with their attributes, and the interactions users have with them — views, purchases, plays, ratings. More signal helps, but you can start small.
How quickly do recommendations react?
Interactions are applied in real time, so the next request already reflects what the user just did.
Can we use it alongside our own models?
Yes. Use vecsai as the primary engine, or as a candidate source that your own ranking layer re-orders.
Works with
The rest of the platform
vecsai DB
Text, vector and hybrid search in one fast engine.
Learn morevecsai Search
Search that understands what people mean.
Learn morevecsai Sim
See the effect of a change before your users do.
Learn moreSee vecsai on your own catalogue
Tell us about your platform and we will show you search, recommendations and simulation running on your data.